RCCA Copilot 8D · 5-WHY · ISHIKAWA / 6M — DRAFTING AID

A 60-second walkthrough · pre-baked, no agent call

How RCCA Copilot keeps the engineer the decider.

The same flow you can run yourself on a sample — shown end to end on one investigation, so you can see how the agent drafts and cites while the human edits, signs, and stays accountable. Try it on a real sample →

1 · The nonconformance

NC-TOUR-001 · Anodize coating thickness below spec floor
Part
PN-AL-3382
Lot
LOT-AN-42
Process
Type II sulfuric anodize
Spec
SPEC-AN-7
Measured
0.4 mil
Required
0.7–1.2 mil

2 · The agent's cited 8D draft

Every section is drafted from tool-retrieved evidence — and the engineer can edit any of it.

D2
step
Problem description Approved

Type II sulfuric anodic coating on lot LOT-AN-42 (PN-AL-3382) measured 0.4 mil, below the SPEC-AN-7 §4.2 floor of 0.7 mil (required 0.7–1.2 mil) — ~43% under the minimum. Engineer note: confirmed against the calibrated reference standard; 3 of 5 points below floor.

D3
step
Interim containment Approved

Quarantine lot LOT-AN-42 and 100% eddy-current inspect at five points per SPEC-AN-7 §4.2; hold downstream lots on the same rectifier pending the voltage finding.

D4
step
Root cause Approved

Bath voltage sagged from the ~15 V setpoint to ~11 V mid-run (PD-AN-42-03/04), dropping current density to 7.9 asf (PD-AN-42-07). Coating growth is charge-driven, so the sustained low current density is mechanistically consistent with the 0.4 mil result, and mirrors recurrence NC-PRIOR-101. Underlying reason for the sag is not yet established.

D5
step
Permanent corrective action Approved

Pending root-cause confirmation: retrieve rectifier output/maintenance logs for the run; if rectifier-confirmed, repair/replace the failing component (as in NC-PRIOR-101) and verify output under load before returning the line to service.

D7
step
Prevent recurrence Approved

Add in-run SPC/interlock on bath voltage and current density with auto-alarm; verify per-run rectifier output checks; require recorded five-point thickness per SPEC-AN-7 §4.2.

3 · Approve the evidence, not the prose

D4 · candidate causes & cited evidence
Machine High confidence

Rectifier output sag took bath voltage from ~15 V to ~11 V, pulling current density to 7.9 asf and starving anodic coating growth — producing the 0.4 mil result. Mechanism matches recurrence NC-PRIOR-101.

PD-LOT-AN-42 · PD-AN-42-03 — bath_voltage_V 11.3 below setpoint PD-LOT-AN-42 · PD-AN-42-07 — current_density_asf 7.9 (low) NC-PRIOR-101 · root_cause — Rectifier B output sagged, lowering current density and starving growth
Environment Low confidence

Bath-temperature excursion as a contributor — excluded by evidence; bath temp was in range.

PD-LOT-AN-42 · PD-AN-42-09 — bath_temp_F 70.0 (in range)
Measurement Insufficient evidence

False-low eddy-current reading (probe calibration or single- vs five-point). No calibration record or full five-point dataset was retrieved.

No supporting evidence — investigate further.

4 · The human signs off

Nothing is CAPA-ready until every section is approved — the gate is computed, never forced.

CAPA READINESS
5/5 SECTIONS SIGNED Export 8D / CAPA ↧

5 · The audit trail is the evidence of human judgment

D2 — Problem description Edited
Agent proposed
Type II sulfuric anodic coating on lot LOT-AN-42 (PN-AL-3382) measured 0.4 mil, below the SPEC-AN-7 §4.2 floor of 0.7 mil (required 0.7–1.2 mil) — ~43% under the minimum.
Engineer's final · signed H. Garza (Quality Eng.)
Type II sulfuric anodic coating on lot LOT-AN-42 (PN-AL-3382) measured 0.4 mil, below the SPEC-AN-7 §4.2 floor of 0.7 mil (required 0.7–1.2 mil) — ~43% under the minimum. Engineer note: confirmed against the calibrated reference standard; 3 of 5 points below floor.

See the full audit trail (all sections + event log) →

6 · Export the signed 8D / CAPA

# 8D / CAPA — NC-TOUR-001: Anodize coating thickness below spec floor

- **Part:** PN-AL-3382
- **Lot:** LOT-AN-42
- **Process:** Type II sulfuric anodize
- **Spec violated:** SPEC-AN-7
- **Measured vs. required:** 0.4 mil vs. 0.7–1.2 mil
- **CAPA-ready:** YES

## D2 — Problem description

Type II sulfuric anodic coating on lot LOT-AN-42 (PN-AL-3382) measured 0.4 mil, below the SPEC-AN-7 §4.2 floor of 0.7 mil (required 0.7–1.2 mil) — ~43% under the minimum. Engineer note: confirmed against the calibrated reference standard; 3 of 5 points below floor.

*Approved by H. Garza (Quality Eng.) on 2026-06-07 23:44 UTC.*

## D3 — Interim containment

Quarantine lot LOT-AN-42 and 100% eddy-current inspect at five points per SPEC-AN-7 §4.2; hold downstream lots on the same rectifier pending the voltage finding.

*Approved by H. Garza (Quality Eng.) on 2026-06-07 23:44 UTC.*

## D4 — Root cause

Bath voltage sagged from the ~15 V setpoint to ~11 V mid-run (PD-AN-42-03/04), dropping current density to 7.9 asf (PD-AN-42-07). Coating growth is charge-driven, so the sustained low current density is mechanistically consistent with the 0.4 mil result, and mirrors recurrence NC-PRIOR-101. Underlying reason for the sag is not yet established.

**Candidate causes (Ishikawa / 6M):**

- **Machine** (High confidence): Rectifier output sag took bath voltage from ~15 V to ~11 V, pulling current density to 7.9 asf and starving anodic coating growth — producing the 0.4 mil result. Mechanism matches recurrence NC-PRIOR-101.
  - Evidence: `PD-LOT-AN-42` · PD-AN-42-03 — bath_voltage_V 11.3 below setpoint
  - Evidence: `PD-LOT-AN-42` · PD-AN-42-07 — current_density_asf 7.9 (low)
  - Evidence: `NC-PRIOR-101` · root_cause — Rectifier B output sagged, lowering current density and starving growth
- **Environment** (Low confidence): Bath-temperature excursion as a contributor — excluded by evidence; bath temp was in range.
  - Evidence: `PD-LOT-AN-42` · PD-AN-42-09 — bath_temp_F 70.0 (in range)
- **Measurement** (Insufficient evidence — investigate further): False-low eddy-current reading (probe calibration or single- vs five-point). No calibration record or full five-point dataset was retrieved.

*Approved by H. Garza (Quality Eng.) on 2026-06-07 23:44 UTC.*

## D5 — Permanent corrective action

Pending root-cause confirmation: retrieve rectifier output/maintenance logs for the run; if rectifier-confirmed, repair/replace the failing component (as in NC-PRIOR-101) and verify output under load before returning the line to service.

*Approved by H. Garza (Quality Eng.) on 2026-06-07 23:44 UTC.*

## D7 — Prevent recurrence

Add in-run SPC/interlock on bath voltage and current density with auto-alarm; verify per-run rectifier output checks; require recorded five-point thickness per SPEC-AN-7 §4.2.

*Approved by H. Garza (Quality Eng.) on 2026-06-07 23:44 UTC.*

---
_Simulated data for demonstration only. Entirely fictional — not affiliated with any employer and not derived from any real nonconformance, process, or specification._
Try it yourself on a real sample →